June 05, 2007
Financial Openness around the World
What do our indicators tell us?
Figure 1: Patterns in Capital Account Openness in 2005. Source: Chinn and Ito, "A New Measure of Financial Openness" (2007) [pdf].
Globalization is a phenomenon often documented by way of anecdote. It's also often documented by way of referring to statistics. For trade one can recount trends in the ratio of exports plus imports to GDP. For financial integration, one can look to how large cross-border transactions are, either in flow or stock terms. The latter is the approach undertaken by Lane and Milesi-Ferretti (e.g., here and here), using the cross-country data base they compiled.
Another way of examining financial integration is by way of tabulating the restrictions on capital account transactions. There are a variety of indicators that take the de jure restrictions and code them as either a continuous variable (Quinn) or as a binary variable ("on" or "off"). I, along with Hiro Ito at PSU, have constructed a data set based upon a coding of the IMF's Annual Report on Exchange Arrangements and Exchange Restrictions, hereafter AREAER.
Figure 1 depicts financial openness, as measured by our index, country by country in 2005. The top 25th percentile in openness are depicted in dark red, while the lowest 25th percentile in lightest pink.
This index is the first principal component of four categories of restrictions.
- The existence of multiple exchange rates (k1)
- Restrictions on current account transactions(k2)
- Restrictions on capital account transactions (k3)
- requirement of the surrender of export proceeds (k4)
While one's primary interest in developing a measure of capital openness, the indicator variable on k3 is a very poor indicator on its own of the openness of the capital account. It's a binary variable, which does not indicate on its own the stringency of restrictions of capital controls. Hence, we smooth k3 by using a 5 year trailing moving average. We then use k1-2 and k4 to provide additional information regarding capital account openness, using the presumption that the existence of other restrictions on international transactions will be correlated with how restrictive the capital account regulations are.
We don't argue that our measure is the best. As we discuss in the paper, there are several indicators in circulation. The greatest virtue of our mesure is its wide coverage -- about 180 countries -- and long sample period, starting in 1970.
The behavior of our index appears to conform to priors. Figure 2 indicates that capital account openness across the world has increased over time, albeit slowly (green bars). However, the figure also highlights the variability in capital account liberalization. Industrial countries started from a higher base and have liberalized throughout the past 35 years. Emerging market economies only began liberalization in the 1990s.
Figure 2: Average indices by period subsamples and country groupings. Source: Chinn and Ito, "A New Measure of Financial Openness" (2007) [pdf].
Figure 3 shows the regional variation in liberalization amongst LDCs. One observation is that capital account openness has not trended upward monotonically; Latin America experienced some retrenchment in the wake of the debt crises of the early 1980s. Another observation is that South Asia/Middle East/Africa has been the slowest to open up. In contrast, Eastern Europe has undertaken the most rapid dismantling of restrictions.
Figure 3: Average indices by period subsamples and LDC regional groupings. Source: Chinn and Ito, "A New Measure of Financial Openness" (2007) [pdf].
So when you hear the conclusions of a model that assumes that capital is mobile, check to see which countries look likely to fit that characterization.
The index has been used in the following papers. Chinn and Ito, JDE 2006 [pdf] on financial development; Cheung, Chinn and Fujii, JIMF f'coming [pdf] on the RMB; Chinn and Ito, JIMF f'coming [pdf] on the saving glut; Joyce and Noy, RIE f'coming [pdf] on the IMF and capital liberalization; Mendoza, Quadrini and Rios-Rull [pdf] on imbalances; di Giovanni and Shambaugh (2005) [pdf] on interest rate links; Eichengreen and Leblang (2007) [pdf] on democracy and globalization; Eichengreen and Razo-Garcia EP (2006) [pdf] on exchange rate regimes; Brooks, World Politics (2004) [pdf] on determinants of liberalization.
Posted by Menzie Chinn at June 5, 2007 06:27 AMdigg this | reddit
While the index appears to be useful there are some concerns about its theoretical underpinnings as traditionnal trade restrictiveness indexes...
Do you think it would be possible to duplicate Kee, Nicita et al.'s methodology, as developed in a recent paper?
Posted by: Bouveret at June 6, 2007 01:51 AM
Although you do not state it specifically, your model seems to be another study making the case that more financial openness is linked to higher standards of living, per capita income, or other variables of well being. A quick glance at the map, to me, correlates very well with the Heritage Foundation's annual ranking of economic freedom.
But, how truly happy are people living in socialist countries as opposed to capitalist countries? You cannot measure happiness by material wealth alone. Blowing oneself up and taking some Jews with you is the absolute pinnacle of happiness for many in the Arab world. Lets be careful we do not force our values on others. Morality is relative. Many people around the world might just be happier in poverty. Many regimes are dependent on it. Encouraging financial openness might just create political instability in such places, encouraging people to revolt against socialist dictators and begin pursuing material wealth and decadence when all they need to be happy is some infidels to kill.
Posted by: Hitchhiker at June 6, 2007 09:05 AM
your analysis of the Arab world seems very simple... I would rather say that they (People from Palestin, not people from Morocco, Tunisia, Algeria, Lybia, Saudi Arabia, Qatar, UAE...) want their country to be independent and well established. Their main goal is to achieve this freedom, sometimes with terrorist means, but i don't think they are happy blowing themselves (i'm not sure they're happy at all either). I think there is a confusion between primary and secondary objectives. Anyway you have a point on the definition of well-being throughout the world. Concerning morality, it's not that relative (read the writings of Immanuel Kant, 18th century for instance), anyway that's another story...
Posted by: Bouveret at June 7, 2007 12:59 AM
Bouveret, I like your work on the limitations of equilibrium exchange rate theories, and I want to thank you for the reference above because I did not know that non-tariff barriers were so important (contributing more than 70% on average to world protection).
Also, see this paper by Dakhlia and Temimi which observes that the trade restrictiveness index may not exist or be unique when countries are large (which they adapt for):
Posted by: Charlie Stromeyer at June 7, 2007 07:47 AM
thanks for the reference too, i'll take a look at it soon.
On equilibrium exchange rates i wrote a previous paper two year ago that may be more documented. Unfortunately the paper is in french, maybe i'll translate it soon...You can find an english summary and the french version at:
Posted by: Bouveret at June 8, 2007 01:53 AM